Understanding individual differences within large-scale brain networks across cognitive contexts 

Historically, human neuroimaging has studied brain regions “activated" during behavior and how they differ between groups of people. This approach has improved our understanding of healthy and disordered brain function, but has two key shortcomings. First, focusing on brain activation restricts how we understand the brain, ignoring vital, behind-the-scenes processing. In the past decade, the focus has shifted to communication between brain regions, or connectivity, revealing networks that exhibit subtle, consistent differences across behaviors and diagnoses. Second, focusing on group differences ignores substantial within-group heterogeneity and often imposes false dichotomies. Recent findings show that brain network variability within an individual is nearly as great as across a group. Altogether, this illustrates a need for understanding individual variability in brain networks and how it relates to behavior. Therefore, I have developed a pipeline for investigating individual differences in brain connectivity, adapting robust statistical methods to address unique challenges of neuroimaging data analysis. Here, I describe this pipeline and apply it to two datasets. First, I explore between-individual variability in brain connectivity underlying intelligence and academic performance to better understand factors contributing to student success. Second, I assess the relative contributions of stress, sleep, and hormones to within-individual variability in brain connectivity across the menstrual cycle to illuminate little-studied phenomena affecting the everyday lives of half the population. Finally, I introduce a novel signal processing workflow for cleaning electrophysiological measures of bodily stress and arousal in neuroimaging research. 

Major Professor: Dr. Angela R. Laird